Content uploaded by Christina Beatty
Author content
All content in this area was uploaded by Christina Beatty
Content may be subject to copyright.
THE DIVERSION FROM ‘UNEMPLOYMENT’ TO ‘SICKNESS’ ACROSS BRITISH
REGIONS AND DISTRICTS
Christina Beatty and Stephen Fothergill
Centre for Regional Economic and Social Research,
Sheffield Hallam University
April 2004
1
Abstract
More than 2.5 million non-employed adults of working age in Britain claim sickness-
related benefits, and the numbers have risen steeply over time. The very large
variation in the numbers across districts and regions points strongly to extensive
hidden unemployment, especially in older industrial areas affected by job losses.
This article builds on two previous articles in Regional Studies – one dealing with the
theoretical framework and the other a local case study - to present wholly new
estimates of the scale of the diversion across all parts of the country. It also
questions contemporary perceptions of the UK labour market and the validity of
current approaches to re-engaging sickness claimants with employment.
Unemployment
Sickness
Social Security
Districts
2
Introduction
There is a widely held view among politicians and journalists that the UK’s
unemployment problem is all but solved. The UK labour market is increasingly seen
as being characterised mainly by labour shortages, and residual unemployment is
frequently attributed to failings in individual skills and motivation.
There is little doubt that the UK labour market has indeed improved substantially
since the depths of the early 1990s recession, and that parts of southern England are
now at or near full employment. However, the assertion that unemployment has all
but faded away is based on serious misunderstandings about what has actually
happened in the labour market. In particular, the denial of continuing and large-scale
joblessness relies all too often on data that measures only part of the overall
problem.
This article explores what is probably the largest single distortion to the data – the
diversion from unemployment to sickness benefits. In particular, the article presents
new and up-to-date estimates of the scale of the diversion and, for the first time in a
journal article, provides estimates of the size of the diversion in every region and
district of Great Britain.
The first part of the article explains how the UK benefits system works and how this
gives rise to a diversion from recorded unemployment to recorded sickness. This is
followed by a review of the existing evidence. The main body of the article then
presents estimates of the scale of the diversion, deploying a new and improved
method that provides more robust estimates at the local and regional scale. The final
part of the article comments on the nature of this form of unemployment, on the
3
implications for perceptions of the contemporary UK labour market, and on the
policies likely to move the British economy closer to genuine full employment.
Two benefit systems
It is not widely recognised that in the UK two separate benefit systems provide
support to non-employed adults of working age.
The first relates to ‘unemployment’. Since 1996 this has taken the form of
Jobseeker’s Allowance (JSA). To claim JSA a person must demonstrate that they
are available for work and looking for work, and they must ‘sign on’ once every two
weeks. For most claimants, including all those claiming for more than six months,
JSA is means-tested (on the basis of household income).
The other benefit system relates to ‘sickness’. Since 1995 this has taken the form of
Incapacity Benefit (IB). IB is paid to non-employed adults of working age who have
health problems or disabilities. About two-thirds of IB claimants actually receive
Incapacity Benefit. The remaining third, with insufficient National Insurance (NI)
credits to qualify for IB itself, are counted as ‘NI credits only’ claimants and in most
cases actually receive means-tested Income Support with a disability premium.
Importantly, Incapacity Benefit is not means-tested except for a small number of new
claimants with substantial pension income. Also, although Incapacity Benefit
payments start at almost the same rate as JSA they increase after six months and
again after twelve months. The disability premium payable to ‘NI credits only’ IB
claimants also makes this worth more than JSA.
The workings of the benefits system may seem a long way removed from the
measurement of unemployment. The point is however that for many of the longer-
term jobless who have health problems, the differential in benefit payments creates
an incentive to claim IB rather than JSA. For example, an unemployed man with a
wife in work and perhaps a small pension from a previous employer will not generally
be entitled to means-tested JSA. In essence, his wife’s earnings and his pension
reduce or eliminate his JSA entitlement. But if he has sufficient health problems, and
if he has enough NI credits (which most men with a work history do have), he will be
eligible to claim Incapacity Benefit irrespective of his wife’s earnings or in most
circumstances of his pension as well.
4
The gatekeepers determining access to Incapacity Benefit are medical practitioners –
initially the claimant’s own GP but for claims beyond six months doctors working on
behalf of the Benefits Agency. To qualify for IB a person must be deemed not fit
enough to work. In practice, however, the tests applied by the Benefits Agency
assess ability to undertake certain basic physical tasks rather than inability to do all
kinds of work in all circumstances. Many unemployed people have picked up injuries
over the course of their working life, and there is the effect on health and physical
abilities of simply getting older. In practice, therefore, many of the unemployed with
health problems are able to claim IB rather than JSA. As IB claimants they are not
required to sign-on fortnightly or to look for work. Instead they will typically be
recalled for medical re-assessment only once every two or three years.
The diversion onto Incapacity Benefit distorts both official measures of UK
unemployment. The best known of these measures is the claimant count – the
number of people out of work and claiming unemployment related benefits, mainly
JSA but also NI credits for unemployment. In the UK benefits system, the men and
women claiming unemployment-related benefits and those claiming sickness-related
benefits are two mutually exclusive groups. None of the Incapacity Benefit claimants
are therefore included in the claimant count.
The other measure of unemployment (and officially the preferred one, even though it
is less often quoted) is the ILO measure derived from the Labour Force Survey. This
uses the International Labour Organisation definition of unemployment which counts
anyone who is out of work and wants a job, is available to start in the next two
weeks, and has looked for work in the last four weeks. The ILO definition produces
unemployment figures for Britain as a whole that in the last three or four years have
been around half a million higher than the claimant count. In theory the ILO measure
of unemployment is independent of benefit rules. In practice, because there is no
requirement for IB claimants to look for work and because many think that they would
not find suitable work, most IB claimants do not look for work. They therefore fail one
of the ILO unemployment tests and drop out of the ILO unemployment figures as well
as the claimant count.
The numbers claiming Incapacity Benefit are now truly astonishing. Figure 1 shows
the number of men and women of working age (16-64 for men, 16-59 for women)
claiming IB (or its predecessor Invalidity Benefit) for more than six months. The
5
numbers have risen more or less continuously for two decades. In 1981 there were
570,000 men and women in this category. By 2003 the figure had risen to 2,130,000.
Even this is not the full picture. Added to this there were more than 300,000 further
claimants of working age receiving Severe Disablement Allowance (SDA), which is
paid to people with a high degree of disability but insufficient NI credits to qualify for
IB. There were also more than 200,000 short-term (ie. less than six months) IB
claimants of working age. Official statistics show that in Britain as a whole in August
2003 a grand total of almost 2.7 million non-employed people of working age were
claiming sickness-related benefits. Of these, 1.6 million were men and 1.1 million
were women.
The big increase in the number of working-age men claiming sickness-related
benefits was primarily a phenomenon of the 1980s and early 1990s. Thereafter, the
number of long-term male claimants shown in Figure 1 tended to plateau with only a
very small continuing increase. 1995 is a significant date in this regard in that it
marks the changeover from Invalidity Benefit to Incapacity Benefit, which transferred
responsibility for authorising longer-term claims from family doctors to doctors
working on behalf of the Benefits Agency and introduced standardised assessment
procedures. The number of women who are long-term IB claimants has in contrast
shown an almost continuous increase, though the absolute level remains lower than
for men. What the trends in Figure 1 also demonstrate is that the long period of
economic growth in the British economy from around 1993 onwards made absolutely
no dent in the number of long-term IB claimants.
It is highly unlikely that there has been a four-fold increase in the level of long-term
incapacitating illness in the UK workforce over the last twenty years. Indeed, the
increase has happened at a time when general standards of health are known to be
showing a slow but steady improvement, admittedly with the slowest improvement
among the most disadvantaged groups. Almost certainly, what can be observed in
the rise in the number of long-term sickness claimants is to a great extent the
interaction of a difficult labour market and the UK social security system.
Existing evidence
That the claimant count measure of UK unemployment is flawed is no longer
disputed. Whilst it is accurate in counting those who are out of work and in receipt of
6
unemployment-related benefits, what is accepted is that the claimant count is
influenced by changes in benefit rules. There have been more than thirty of these
since the early 1980s, not least the changeover from Unemployment Benefit to
Jobseeker’s Allowance, the effect of which was to cut the claimant count by reducing
the duration for which non-means tested benefits were available and increasing the
requirement to demonstrate active job-seeking. The criticisms of the claimant count
have come from academic sources (for example Gregg 1994, MacKay 1999), from
independent watchdogs such as the Unemployment Unit (Convery 1996) and from
no less a source than the Royal Statistical Society (1995).
A study of the labour market in the UK coalfields in the wake of pit closures was one
of the first to argue that there is a diversion from unemployment to sickness benefits
(Beatty and Fothergill 1996). The study found that the largest single adjustment to
job loss was a withdrawal of men into ‘economic inactivity’, and that the largest
component of this withdrawal was a big increase in recorded ‘permanent sickness’.
In contrast, pit closures had virtually no impact at all on recorded unemployment in
the coalfields. Subsequent similar studies of England’s disadvantaged rural areas
(Beatty and Fothergill 1997) and of seaside towns (Beatty and Fothergill 2003) also
identified withdrawals from the labour market into ‘sickness’ as the key factor holding
down recorded unemployment among men, though not on quite the same scale as in
the coalfields. A study of Britain’s cities (Turok and Edge 1999) likewise found that
labour market withdrawal by men was a key response to job loss in the 1980s and
early 1990s.
More generally, there is accumulating evidence that the disparities in employment
opportunities between different parts of Britain are reflected less in unemployment
data than in levels of economic inactivity among the working age population. Studies
by Gregg and Wadsworth (1998), Green (1997,1999) and Green and Owen (1998)
confirm this point. Mackay (1999) makes the observation that “the greater the
degree of labour market disadvantage, the less appropriate is unemployment as a
measure of labour market slack”. The relevant point here is that whereas the
unemployed (either claimant or ILO) are conventionally included among the
economically active, claimants of sickness-related benefits such as Incapacity Benefit
are nearly all included among the inactive.
The relationship between unemployment, ill health and the number claiming
sickness-related benefits is nevertheless complex. In parallel with the argument that
7
there has been a diversion from ‘unemployment’ to ‘sickness’ within the benefits
system, there is a quite separate argument that unemployment is actually a cause of
ill health. That there is a causal link from unemployment and poverty to ill health is in
fact not disputed (see for example Bellaby and Bellaby 1999). Nor is the fact that the
groups most at risk of unemployment are also the ones most likely to be affected by
ill health (Bartley and Owen 1996). But these processes alone seem unable to
account for the sheer number of sickness claimants in the UK, or the scale of the
increase through time. The UK’s General Household Survey, for example, records
an increase of nearly a quarter in self-reported limiting long-term illness among men
between the early 1980s and the late 1990s, but this increase is of a magnitude that
seems unable to explain the huge increase in the number of sickness claimants.
Moreover, General Household Survey data may reflect a greater willingness over
time to report illnesses as much as any deterioration in underlying standards of
health.
Beatty, Fothergill and Macmillan (2000) tried to reconcile the rising numbers claiming
Incapacity Benefit with the observation that standards of health have not deteriorated
to the same extent. They also tried to reconcile the rising numbers with the
requirement for all IB claimants to demonstrate a significant degree of ill heath. The
argument put forward was that work-limiting ill health is actually quite widespread in
the working age population but that many of the men and women with health
problems do in fact hold down jobs. This is confirmed by data from the Labour Force
Survey (Labour Market Trends 2002) showing that of the estimated 7.2 million men
and women of working age in the UK who had a work-limiting long-term illness or
disability in winter 2001/02, 3.4 million were in employment.
Beatty, Fothergill and Macmillan argued that difficult labour market conditions such
as those experienced in the UK for most of the 1980s and 90s expose men and
women with health problems to job loss, and that when they find themselves out of
work their health often places them at the back of the queue for jobs. These people
have sufficient ill health to access sickness-related benefits instead of unemployment
benefits – in other words their benefit claims are not fraudulent. The net effect,
however, is that in a difficult labour market the ill heath that was once hidden
because people were in work becomes visible in the numbers claiming sickness-
related benefits. Furthermore, as the economy picks up it is the healthy job-seekers
on unemployment benefits who are the first to be taken on again, leaving a large
8
marginalized group on sickness benefits – precisely the experience in the UK labour
market in the late 1990s and early 2000s.
The survey evidence on Incapacity Benefit claimants tends to confirm this
perspective. Easterlow and Smith (2003) found that people experiencing ill health do
not lack the incentive to work and are likely to be forced rather than lured onto
pensions and onto benefits. Extensive survey work reported in Alcock et al (2003)
confirms that although a degree of self-reported work-limiting ill health is just about
universal among men claiming Incapacity Benefit, only a quarter say they can’t do
any work at all. The same survey evidence also shows that active job-seeking tails
off badly among male Incapacity Benefit claimants, with barely one in twenty looking
for work despite the fact that around half say they would like a full-time job and more
than a quarter looked for work after their last job ended.
A survey of male Incapacity Benefit claimants in Barrow-in-Furness (Beatty and
Fothergill 2002a), an area affected by major job losses in the local shipbuilding
industry, provides further evidence that a substantial proportion of IB claimants
should be regarded as hidden unemployed. In Barrow the age, skills and low
qualifications of male IB claimants would anyway have exposed them to
unemployment. Ill health was the reason for job loss in fewer than half of all cases,
with redundancy often figuring strongly. Two-thirds of Barrow’s male IB claimants
said they would like a full-time job, and only a third said they could do no work at all.
Analyses of local and sectoral data offer additional evidence. Armstrong (1999)
concluded that there is evidence of hidden male unemployment among sickness
claimants in Northern Ireland. In North West England, Sutherland (1999) highlighted
off-flows from claimant unemployment onto Incapacity Benefit. Fieldhouse and
Hollywood (1999) found a strong movement of ex-miners into permanent sickness
rather than recorded unemployment.
The geography of sickness benefit claimants
What is particularly striking is the distribution of sickness claimants across Britain.
Figures 2 and 3 illustrate this point. They show the share of the total working age
population, by district in August 2003, claiming sickness-related benefits – in this
instance Incapacity Benefit (long and short-term and NI credits only) and Severe
9
Disablement Allowance. The headline GB total at that time was 2,662,000,
representing 7.5 per cent of the entire working age population. The data on the
number of claimants in each district comes from the Department for Work and
Pensions and is based on a 5 per cent sample of claimants, which in view of the
exceptionally large numbers on these benefits can be considered to provide a
reliable picture. The number of claimants in each district is expressed as a
percentage of the 2002 working age population, again from official data.
What is immediately apparent is that sickness claimants are especially concentrated
in certain areas, notably North East England, Merseyside, South Wales, parts of
Yorkshire, and Clydeside. These are the parts of Britain where industrial job losses
have been concentrated over many years and where claimant unemployment has
persistently been higher than the national average.
Table 1 lists the top 20 and bottom 10 districts across Britain in terms of the share of
the working age population claiming sickness-related benefits. The top 20 are
without exception older industrial districts in the North, Scotland and Wales. Around
half are former coalmining areas and no fewer than seven cover the Welsh Valleys.
The top 20 also includes some substantial cities – Glasgow, Liverpool, Stoke and
Manchester. Inner London boroughs, which often have relatively high claimant
unemployment, are conspicuous by their absence from this list. At the other end of
the scale, the bottom 10 are all small town and rural districts in the south and east of
Britain. The difference between the extremes is considerable – there are twelve
times as many sickness claimants, in relation to the local population, in Easington
district in County Durham as in Hart district in Hampshire.
In Easington and in Merthyr Tydfil in the Welsh Valleys, more than one in five of all
adults of working age – ie of all 16-64 year old men and all 16-59 year old women -
are out of work and on sickness benefits. In Glasgow and in Liverpool the proportion
is one in six. In Glasgow, 63,600 men and women of working age are on these
benefits. In Liverpool the figure is 45,000. The proportions among men are still
worse: in Easington and Merthyr Tydfil, 24 per cent of all working age men – just
under one in four – are out of work and claiming sickness benefits. None of these
men are included in the claimant unemployment figures.
Within specific segments of the workforce the incidence of sickness claimants is even
greater. Men are more likely than women to be sickness claimants, but within the
10
male workforce it is older, manual workers with few formal qualifications who are
most likely to claim these benefits (see Alcock et al 2003). The likelihood of claiming
sickness benefits rises sharply with age, which is consistent with the view that older
workers are more likely to experience the health problems that enable them to claim
Incapacity Benefit rather than Jobseeker’s Allowance. In Merthyr Tydfil, admittedly
an extreme case, just over half of all men aged between 50 and 64 were sickness
claimants in May 2002 (Beatty and Fothergill 2002b). In Glasgow the proportion was
44 per cent, in Liverpool 38 per cent and in Manchester 36 per cent.
In total there are 68 districts in England, Scotland and Wales where in August 2003
ten per cent or more of the entire working age population was out of work and
claiming sickness-related benefits. Not a single one of these districts was in London,
the South East, South West or Eastern England.
Measuring the diversion
The crucial issue is the extent to which these huge numbers represent hidden
unemployment. The key analytical problem is that the headline figures for IB
claimants conflate two groups – those whose health problems are so severe that they
would remain on sickness benefits in all circumstances, and those who would have
been in work if suitable jobs had been available. At the level of the individual, in
particular, the dividing line is not easy to draw. Additionally, there are differences
between areas in the underlying health of the population which mean that the number
of sickness claimants is always going to vary from place to place.
Partly the problem is one of definition. In this article we set out to measure the
number of sickness claimants who could reasonably be expected to have been in
work in a fully employed economy. This is not the same as the number who are
actively looking for work. Nor is it the same as the number who have actually moved
directly from unemployment-related benefits (mainly JSA) to sickness-related
benefits (mainly IB) since large numbers move directly from employment onto
sickness benefits.
The basic approach adopted here is to establish a ‘benchmark’ reflecting what is
achievable in a fully employed economy, and to compare actual levels in each district
with that benchmark. Levels above the benchmark are deemed to represent hidden
11
unemployment. In principle this is the method used in earlier studies (for example
Beatty and Fothergill 1996) but here an improved and more robust version is
deployed.
There are two components to the benchmark used here. The first is the proportion of
men and women of working age who are sickness claimants in fully-employed parts
of Britain. The area chosen here to represent a ‘fully-employed economy’ comprises
the seven counties of Berkshire, Buckinghamshire, Hampshire (minus Portsmouth
and Southampton), Hertfordshire, Oxfordshire, Surrey and West Sussex. These
make up a block to the north, west and south of London where by 2003 there had
effectively been full employment for four or five years. The share of the working-age
population in work – the ‘employment rate’ – averaged in excess of 80 per cent in
each of these counties in 2003. The employment rate in neighbouring counties to the
east and south east of London (Essex, Kent, East Sussex) fell a little below this 80
per cent threshold.
In August 2003, 4.1 per cent of working age men and 3.3 per cent of working age
women were sickness benefit claimants in this fully employed part of southern
England. This low level illustrates what can be achieved in contemporary Britain, at
least in some areas, in the context of full employment.
The second component of the benchmark is the underlying deviation in sickness
levels in each district from the level in this fully employed part of southern England.
Here, as a guide, we use the proportion of men and women of working age who were
recorded as ‘permanently sick’ by the 1981 Census of Population, when the figures
were still largely unaffected by the subsequent diversion into hidden unemployment1.
This is the approach first deployed by Armstrong (1999). The excess in the
proportion of ‘permanently sick’ in each district in 1981, over the comparable 1981
figure for the fully-employed part of the South, is added to the sickness claimant rate
in the fully-employed part of the South in 2003 to give an overall benchmark for the
district. The higher underlying level of incapacitating ill health in some areas is
therefore built into the benchmark.
In a small number of mainly rural districts the data on permanent sickness in 1981 is inflated
by the location of large psychiatric institutions, which have virtually all subsequently closed.
To adjust for this potential distortion, in the districts where according to the 1981 Census of
Population the proportion of the working age population living in such institutions exceeded
one per cent, the excess is deducted from the 1981 permanent sickness data for that district.
12
1
In each district the benchmark therefore takes account not only of the level of
sickness claimants achievable in a fully-employed part of Britain but also of
geographical variations in underlying ill-health. The benchmarking exercise has been
carried out separately for men and women.
An example will make the procedure clearer. Take the case of men in Barnsley,
shown in Table 2. Barnsley, a former coalmining district in South Yorkshire, has a
total population of around 230,000 and a working age population of just over
130,000. In Barnsley in August 2003, 12,000 men of working age were out of work
and claiming sickness-related benefits. This represented 17.4 per cent of the entire
male working age population. The benchmark for Barnsley, as elsewhere, comprises
two elements. First there is the sickness claimant rate among men in the fully-
employed parts of the South – 4.1 per cent, equivalent to 2,800 men in the Barnsley
context. Second there is the excess sickness among working age men in Barnsley,
over the level in this part of the South, recorded in 1981 before the figures became
badly contaminated by the diversion from unemployment – 3.4 per cent, equivalent to
2,400 men. This excess reflects the poorer underlying standard of health in
Barnsley, not least as a result of employment in the coal industry. These figures give
an overall benchmark of 7.5 per cent, or 5,200 men. This is the number of male
sickness claimants that we would expect in Barnsley in the context of full
employment. The difference between this figure and the actual number of male
sickness claimants – 6,800 or 9.9 per cent of the male working age population - is
what we identify as the diversion from unemployment. For comparison, the number
of men who were claimant unemployed in Barnsley in August 2003 was just 2,217.
Barnsley is a fairly extreme case: it has the eleventh highest male sickness claimant
rate in Britain. The point is however to illustrate the estimation process. Later we
subject the resulting estimates to cross-checking by other methods, but at this point it
is worth noting that the estimation procedure implicitly assumes that the underlying
geography of incapacitating ill health has not changed between 1981 and 2003. In
practice there will of course have been changes, but whether these will have greatly
altered the relativities between districts is questionable bearing in mind the
considerable stability of socio-economic disparities at the district scale that underpin
differences in standards of health. Additionally, even in 1981 there may have been
limited spill-over from unemployment to sickness benefits in some districts. The
figures on the diversion to sickness benefits generated by these methods therefore
need to be treated as estimates. On the other hand they do attempt to take account
13
not only what has been shown to be achievable in fully-employed areas but also of
underlying differences between districts.
New estimates of the diversion from ‘unemployment’ to ‘sickness’
Beatty et al (2002) first applied the method outlined above to data for August 2001.
Here we present new estimates for every district in Great Britain based on sickness
benefit data for August 2003.
Figures 4 and 5 map the scale of estimated diversion from unemployment to
sickness benefits. What the maps show is a profoundly uneven distribution. At one
end of the scale there are thirty districts where there is estimated to be no diversion
at all. These are, in effect, the districts where full employment already prevails and
there is no reason why anyone need remain on Incapacity Benefit if they want to
work and are able to work. These districts are nearly all concentrated in the south
and east of Britain, especially in an area to the west of London.
At the other end of the scale, there are a large number of districts in the older
industrial areas of the North, Scotland and Wales where the diversion from
unemployment appears to be substantial. In a handful, including Glasgow and
Liverpool, the estimated diversion exceeds ten per cent of the entire adult workforce.
By and large, the areas where this large diversion is estimated to occur are the same
districts where the overall share of the working age population claiming sickness
benefits is particularly high. In other words, even after adjusting for differences in
underlying ill health these areas have large numbers of men and women who appear
to have been diverted from unemployment.
Table 3 shows the regional pattern of the estimated diversion from unemployment to
sickness benefits. This confirms the skewed regional distribution evident in the
maps. In the North East, North West, Scotland and Wales, 5-6 per cent of the
population of working age is estimated to have been diverted from unemployment.
The proportion in the South East of England is below 1 per cent, and only 1-2 per
cent in London, the South West and Eastern England. The Midlands and Yorkshire
fall between the two extremes.
14
Table 3 also shows the numbers estimated to be part of this diversion. Across Britain
as a whole it is estimated that 1,130,000 people have been diverted from
unemployment to sickness benefits – 650,000 men and 470,000 women. For
comparison, total claimant unemployment across Britain at the same time (August
2003) stood at just 911,000. The comparison is illuminating: it suggests that Britain
has more ‘hidden’ unemployed among sickness claimants than ‘visible’ unemployed
on the claimant count.
Table 4 compares these national estimates with the total numbers of sickness-related
benefit claimants. This too is illuminating because it suggests that despite the
enormous scale of the apparent diversion from unemployment, it accounts for only
just over 40 per cent of all sickness claimants of working age. This observation holds
for both men and women. In effect, what this comparison is telling us is that even if
genuine full employment could be achieved in every part of the country, the total
number of sickness claimants, currently rather over 2.6 million, could be expected to
fall no lower than about 1.5 million.
Sickness benefit numbers around 1.5 million, even in the context of full employment,
would still represent a substantial increase on historic levels, which in the early 1980s
were well below the million mark. In purely statistical terms what this reflects in our
calculations is the fact that even in the fully-employed parts of the South, sickness
claimant rates are now higher than they were in the early 1980s. A little of the
increase through time may reflect an ageing population of working age – fewer
younger workers and more over 50. The rising number of women in paid
employment, who therefore accrue the NI credits that entitle them to Incapacity
Benefit, may account for a little more of the increase.
More likely, changes in the workplace explain the largest part of the underlying
increase in the number of sickness claimants. Whereas the numbers in physically
strenuous or dangerous industries such as mining has declined, it is often argued
that the pace and pressure in most workplaces has increased. Thus whereas at one
time employers might have moved workers with health problems onto lighter or less
demanding duties, it is questionable whether the scope for this practice now exists on
anything like the same scale. The space for the less healthy worker has often gone,
and sickness that was once hidden in the workplace has become visible in the
benefits system.
15
Table 5 shows the top 20 districts in terms of the estimated share of the working age
population diverted from unemployment to sickness benefits. Easington once more
tops this list – one in seven of the entire working age population is estimated to fall
into this category. The top 20 again includes Glasgow and Liverpool. Middlesbrough
and Stoke-on-Trent also figure on the list. In these top 20 districts alone, a total of
more than 210,000 people are estimated to have been diverted from unemployment
to sickness benefits.
Finally, Table 6 shows the scale of the estimated diversion in each of Britain’s 28
principal cities. The range varies from 11.2 per cent of the working age population in
Glasgow to 1.9 per cent in Portsmouth, illustrating the point that a large-scale
diversion is not something that is common to all Britain’s cities. Rather, it is the
former industrial cities of the North, Scotland and Wales that are estimated to have
experienced the largest diversion. Although the absolute numbers in London are
large, it comes third from bottom on the list in terms of share of the working age
population.
How reliable?
The estimates of the number diverted from unemployment to sickness benefits (or
more particularly the men within this group) can be cross-checked against estimates
derived by four alternative methods. Two are statistical comparisons using
alternative benchmarks. One uses South East sickness data for 1991, when this
region had just come to the end of an earlier period of effectively full employment.
This 1991 South East benchmark was the basis of earlier published estimates of
hidden unemployment (for example Beatty and Fothergill 1996). The other statistical
comparison uses national sickness data for 1981, before the figures became badly
contaminated by the diversion from unemployment2. The third and fourth alternative
estimates are based on survey data for male IB claimants (see Alcock et al 2003).
In Britain as a whole in August 2003 a total of 1,470,000 men of working age were
claiming Incapacity Benefit. The four alternative methods generate the following
estimates of hidden unemployment among this group of men across Britain as a
whole:
Both statistical comparisons deduct short -term (ie less than six months) IB claimants from
the headline figure to improve comparability with Census ‘permanent sickness’ data.
16
2
Using the level of ‘permanent sickness’ 680,000
in the South East in 1991 as the benchmark
Using the national (GB) level of ‘permanent 730,000
sickness’ in 1981 as the benchmark
Using the share of male IB claimants 690,000
who say they would like a full-time job (47%)
Using the share of male IB claimants 760,000
who lost their last job mainly for reasons
other than ill-health or injury (52%)
The number of men diverted from unemployment to sickness benefits across Britain,
generated by the methods used in this article, is 650,000. The fact that five separate
methods point to a diversion of between 650,000 and 760,000 gives considerable
confidence. The method adopted in this article, however, remains the one most likely
to generate robust figures at the district scale because it takes account not only of
what has already been shown to be possible in fully-employed areas but also of
underlying geographical variations in incapacitating ill health
A further check on the reliability of the estimates comes from a comparison with
Labour Force Survey data. An analysis of this information for 2001-02 (Labour
Market Trends 2002) shows that of the 7.2 million men and women of working age in
the UK who had a current work-limiting health problem or disability, 15.7 per cent or
1.13 million were economically inactive but said they would like a job. By
comparison, the total number of men and women in Great Britain in August 2003 who
are estimated to have been diverted from unemployment to sickness benefits using
the methods in this report is identical at 1.13 million, though our methods point to
around 50,000 more men and 50,000 fewer women than this comparison with LFS
disability data.
There must nevertheless be a residual question mark about the estimates for
women. The rising number of women claiming sickness-related benefits sits oddly
alongside what is generally recognised to be a labour market that is providing
17
growing job opportunities for women. At least part of the increase in the number of
women claiming these benefits may in fact represent a diversion from other benefits
(such as Income Support for lone parents) rather than from unemployment. Exactly
how many of the 1.1 million women of working age on sickness benefits might have
been in work in a fully employed economy is a question that really requires further
research.
In what sense unemployed?
It is important to be clear about the nature of the diversion from unemployment to
sickness benefits that our methods identify. These are people who might reasonably
be expected to have been in work in a fully employed economy. As we noted earlier,
they are not necessarily active job seekers. However, the fact that some do not
actively look for work should not disqualify them for inclusion because where
appropriate jobs are in short supply many people are realistic enough to know that
they are unlikely to find suitable employment. They therefore give up looking for
work, but that does not make them any less unemployed.
It is also important to emphasise that there is nothing fraudulent about the behaviour
of the large number of people who now claim Incapacity Benefit. All these men and
women will have been given the appropriate medical certification to entitle them to IB.
The health limitations are genuine, even if not necessarily always fully incapacitating
in all circumstances. What has happened is that job loss has fallen
disproportionately on less healthy workers, many of whom are also older and less
skilled. These people have then found themselves at the back of the queue for jobs.
The benefits system and the employment services have then interacted to divert
large numbers away from recorded unemployment and into recorded sickness.
Hidden unemployment, on Incapacity Benefit, is nevertheless different in some ways
from conventional claimant unemployment. Because so many of the men and
women on Incapacity Benefit have given up actively seeking work their
unemployment may be less painful than for JSA claimants. They no longer have to
endure failed job applications and dashed hopes. Also, because of their detachment
from the labour market they do not form part of the stock of potential workers from
whom employers choose and consequently they exert no downward pressure on
wage inflation. Often the hidden unemployed may have become reconciled to their
18
position outside the labour market. In a sense what has happened, after two
decades in which labour markets have been slack in many parts of the country, is
that unemployment has filtered down to rest with the groups who find it most difficult
to hold onto jobs – older workers, the less healthy, less skilled and women with
young children at home. As their status on the margins of the workforce has been
consolidated, these groups have increasingly drawn on benefits other than JSA or
have been denied access to benefits altogether. Their unemployment has slipped
from view.
But none of this changes the fact that very large numbers of who now claim sickness-
related benefits, above all Incapacity Benefit, could have been expected to be in work
in a genuinely fully employed economy. Indeed, the much smaller number of
sickness claimants prior to the 1980s and 90s indicates that far more people with
health problems were once in employment.
Perceptions of the British labour market
The analysis presented in this article represents a challenge to the way that the
contemporary British labour market is perceived. The conventional view, largely
rooted in claimant unemployment data, is that the economy is close to full
employment with residual joblessness on only a modest scale in some parts of the
country.
There is nothing in the evidence presented here that contradicts the view that the
British labour market has improved a great deal during the long period of continuous
economic growth from around 1993 onwards. This has cut the number of claimant
unemployed by around two-thirds. However, the figures presented here show that
economic growth has had no discernible impact on the number of jobless people
claiming sickness benefits. The men and women on these benefits now constitute by
far the largest group of working-age claimants. Furthermore, our analysis indicates
that more than a million of the men and women claiming sickness benefits have been
diverted from unemployment. In other word, the true extent of unemployment is
much greater than official figures indicate.
Labour ministers are therefore entirely correct in arguing that too many people have
become “parked” on Incapacity Benefit. However, what the same ministers have
19
failed to acknowledge is the extent to which this now hides the real level of
unemployment.
The other important conclusion from our analysis is that the regional imbalances in
British labour markets are far more severe than has generally been recognised, and
certainly far worse than claimant unemployment figures suggest. The problem is not
simply that large numbers have been diverted from unemployment to sickness
benefits but that this diversion has happened predominantly in the older industrial
areas of the North, Scotland and Wales. These areas not only have above average
claimant unemployment (which is recognised already) but also exceptionally large
numbers who have been diverted from unemployment to sickness benefits. In
contrast there are substantial parts of the South of England where, even after taking
account of the diversion onto sickness benefits, full employment (or something close
to it) genuinely does appear to exist.
It is undoubtedly true that many of the men on Incapacity Benefit are the workers
who were displaced during the last two decades from industries such as coal, steel
and heavy engineering, and that many of these men are now in their fifties and early
sixties. As they reach state pension they will move off IB. Crucially, however, when
this marginalized cohort finally reaches retirement age they will not free up jobs for
the generations behind them, including the young people continuing to enter the
workforce each year. What this means for many areas is that the present imbalance
in the labour market is likely to be perpetuated. Only job creation (on the labour
demand side) or out-migration and out-commuting (on the labour supply side) would
restore the balance that is currently missing. Furthermore, there is the real prospect
that with tighter rules now controlling access to Incapacity Benefit more of the
continuing imbalances in weaker local labour markets will eventually begin to show
up as claimant unemployment. In other words, the extensive local joblessness
identified by our figures will not just fade away as the large groups of ex-miners, ex-
steelworkers and others finally reach pension age.
Implications for policy
The UK government is well aware of the need to bring down the very large numbers
claiming Incapacity Benefit in order to boost labour supply, raise the capacity of the
economy and cut the benefits bill. Its current approach is a variant of the policies
20
deployed in New Deal programmes for the claimant unemployed. The emphasis is
on the individual – on providing advice and training, re-motivation and the removal of
financial disincentives to return to work. In the case of IB claimants, the measures
are backed up by the intention to introduce more regular medical assessments.
It is hard to be critical of advice and practical support when it is generally welcome to
those who benefit from it. On the other hand, the government’s approach to
Incapacity Benefit claimants can be criticised because the resources made available
to date fall far short of those devoted to unemployed JSA claimants. As a result, the
employment services have so far barely scratched the surface of the Incapacity
Benefit problem, with few claimants benefiting from the emerging initiatives.
A more serious criticism, supported by the data presented here, is that the starting
point of the government’s efforts is simply wrong. The initiatives implicitly assume
that this is a labour supply problem. The marked concentration in Britain’s older
industrial areas, on the other hand, suggests that in fact it is a labour demand issue.
Very large numbers have been diverted from unemployment to sickness benefits in
these areas because there have not been enough suitable jobs in these places.
Indeed, the rising numbers on Incapacity Benefit in these areas represents the
principal labour market response to job destruction in the 1980s and 1990s.
The counter-argument to this criticism is that even if job destruction was the ultimate
cause of the large-scale shift onto Incapacity Benefit, these marginalized workers will
not be re-engaged with the labour market without significant supply-side intervention.
There is probably some truth in this. In practice, substantial numbers of older men
on Incapacity Benefit will probably not now be re-engaged with the labour market in
any circumstances. Their marginalization has become entrenched by years without
work and they can just about get by on a combination of IB, other top-up benefits,
perhaps a small pension from a former employer and possibly a spouse’s income as
well. Some still harbour vague aspirations to work, but others have given up entirely
and now see themselves as retired. In this sense Incapacity Benefit acts as a bridge
to state pension.
The trouble is that labour supply measures are most relevant where there is a strong
demand for labour. Putting aside certain skills that are in short supply everywhere, in
the UK context the areas where there is unquestionably a strong demand for labour
are mostly in the South of England. Yet it is in the South that the diversion from
21
unemployment to sickness benefits appears to be modest at best. There therefore
seems to be only limited scope for boosting labour supply in the South through
targeting IB claimants, though if IB claimants there can be re-engaged with the labour
market they are likely to find a ready demand for their skills.
The much larger problem in the older industrial areas of the North requires a different
approach. Fundamentally, labour demand in these regions needs to be boosted.
This has happened as a result of national economic growth over the last ten years
but, as we explained, the main effect has been to reduce claimant unemployment
leaving the stock of IB claimants largely untouched. Unfortunately, further
macroeconomic stimulus to growth, for example through lower interest rates, would
now run up against the constraint of an already tight labour market in large parts of
southern England. This in turn would risk fuelling inflation.
There is an inexorable logic here that points to regional economic policy as the way
forward. The pressing need is for policies that divert incremental demand for labour
to the parts of the country where there remains substantial labour market slack
among the claimant unemployed but more particularly among the very large numbers
diverted onto sickness benefits. Indeed, with full employment in parts of the South,
regional economic policy is arguably now the essential tool to achieve the
government’s stated goal of full employment.
The rhetoric emerging from Britain’s Labour government is in this respect sending all
the right messages. Regional policy is higher on the agenda than for more than two
decades. However, whether the aspirations are yet matched by practical policies
that can really deliver economic convergence between the UK regions still remains in
doubt.
22
References
Alcock, P., Beatty, C., Fothergill, S., Macmillan, R. and Yeandle, S. (2003) Work to
Welfare: how men become detached from the labour market, CUP, Cambridge.
Armstrong, D. (1999) ‘Hidden male unemployment in Northern Ireland’, Regional
Studies, vol 33, pp 499-512.
Bartley, M. and Owen, C. (1996) ‘Relation between socio-economic status,
employment and health during economic change, 1973-93’, British Medical Journal,
vol 13, pp 445-49.
Beatty, C. and Fothergill, S. (1996) ‘Labour market adjustment in areas of chronic
industrial decline: the case of the UK coalfields’, Regional Studies, vol 30, pp 627-40.
Beatty, C. and Fothergill, S. (1997) Unemployment and the Labour Market in Rural
Development Areas, Rural Research Series no. 30, Rural Development Commission,
London.
Beatty, C. and Fothergill, S. (2002a) ‘Hidden unemployment among men: a case
study’, Regional Studies, vol 36, pp 811-823.
Beatty, C. and Fothergill, S. (2002b) Moving Older People into Jobs: Jobcentre Plus,
New Deal and the job shortfall for the over 50s, Third Age Employment Network,
London.
Beatty, C. and Fothergill, S. (2003) The Seaside Economy, CRESR, Sheffield Hallam
University, Sheffield.
Beatty, C., Fothergill, S., Gore, T. and Green, A. (2002) The Real Level of
Unemployment 2002, CRESR, Sheffield Hallam University, Sheffield.
Beatty, C., Fothergill, S. and Macmillan, R. (2000) ‘A theory of employment,
unemployment and sickness’, Regional Studies, vol 34, pp 617-30.
23
Bellaby, P. and Bellaby, F. (1999) ‘Unemployment and ill-health: local labour markets
and ill-health in Britain 1984-91’, Work, Employment and Society, vol 13, 461-82.
Convery, P. (1996) ‘How many people are unemployed?’, Working Brief, no 78, pp
23-26.
Easterlow, D. and Smith, S. (2003) ‘Health and employment: towards a New Deal’,
Policy and Politics, vol 31, pp 511-33.
Fieldhouse, E. and Hollywood, E. (1999) ‘Life after mining: hidden unemployment
and changing patterns of economic activity among miners in England and Wales
1981-91’, Work, Employment and Society, vol 13, 483-502.
Green, A. E. (1997) ‘Exclusion, unemployment and non-employment’, Regional
Studies, vol 31, pp 505-20.
Green, A. E. (1999) ‘Insights into unemployment and non-employment in Europe
using alternative measures’, Regional Studies, vol 33, pp 453-64.
Green, A. E. and Owen, D. (1998) Where are the Jobless?: changing unemployment
and non-employment in cities and regions, Policy Press, Bristol.
Gregg, P. (1994) ‘Out for the count: a social scientist’s account of unemployment
statistics in the UK’, Journal of the Royal Statistical Society A, vol 157, pp 253-70.
Gregg, P. and Wadsworth, J. (1998) Unemployment and non-employment:
unpacking economic inactivity, Economic Report no 12, Employment Policy Institute,
London.
Labour Market Trends (2002) vol 110, p. 298.
MacKay, R. (1999) ‘Work and nonwork: a more difficult labour market’, Environment
and Planning A, vol 31, pp 487-502.
Royal Statistical Society (1995) Report of the Working Party on the Measurement of
Unemployment in the UK, Royal Statistical Society, London.
24
Sutherland, J. (1999) ‘ Further reflections on hidden unemployment: an examination
of the off-flows from the claimant count in the North West of England’, Regional
Studies, vol 33, pp 456-76.
Turok, I. and Edge, N. (1999) The Jobs Gap in Britain’s Cities, Policy Press, Bristol.
25
Table 1 : Districts with the highest and lowest sickness claimant rates,
August 2003
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
397.
398.
399.
400.
401.
402.
403.
404.
405.
406
% of total
working age
population
Top 20
Easington 21.1
Merthyr Tydfil 20.7
Blaenau Gwent 19.1
Neath Port Talbot 17.2
Glasgow 17.2
Rhondda Cynon Taff 16.7
Liverpool 16.1
Knowsley 16.0
Caerphilly 15.6
Bridgend 14.7
Barrow in Furness 14.4
Wear Valley 14.4
Torfaen 14.4
Barnsley 14.4
Inverclyde 14.2
North Lanarkshire 14.1
Stoke on Trent 14.0
Sedgefield 13.4
Manchester 13.3
Gateshead 13.0
Bottom 10
Surrey Heath 2.8
Wycombe 2.8
Vale of White Horse 2.8
Elmbridge 2.8
South Northamptonshire 2.7
West Berkshire 2.7
Waverley 2.7
South Cambridgeshire 2.7
Wokingham 2.0
Hart 1.7
Sources : DWP and OPCS
Note : Sickness claimants refers to Incapacity Benefit (including NI credits only)
and Severe Disablement Allowance
26
Table 2 : The diversion from unemployment to sickness benefits : a worked
example for men in Barnsley
no. as % male working age
in Barnsley population
Male sickness claimants, Aug 2003 12,000 17.4
BENCHMARK
(1) Male sickness rate in fully-2,800 4.1
employed part of South, Aug 2003
(2) Excess ‘permanent sickness’ 2,400 3.4
in Barnsley over full-employed part
of South, April 1981
Benchmark for Barnsley 5,200 7.5
Estimated diversion 6,800 9.9
(Actual minus benchmark)
Sources : DWP, Census of Population
27
Table 3 : Estimated diversion from unemployment to sickness benefits by
region, August 2003
no. as % total working
age population
North East 95,000 6.1
Wales 101,000 5.8
North West 231,000 5.6
Scotland 172,000 5.4
Yorkshire & Humber 100,000 3.3
West Midlands 107,000 3.3
East Midlands 73,000 2.8
London 103,000 2.1
South West 56,000 1.9
Eastern 47,000 1.4
South East 45,000 0.9
Great Britain 1,130,000 3.2
Source : Authors’ estimates based on DWP data
28
Table 4 : Estimated diversion from unemployment to sickness benefits, by sex,
Great Britain, August 2003
Total no. of sickness Estimated diversion
claimants of working age from unemployment
no. as % total
Men 1,587,000 650,000 41.2
Women 1,075,000 470,000 44.2
Total 2,662,000 1,130,000 42.2
Source : Authors’ estimates based on DWP data
29
Table 5 : Estimated diversion from unemployment to sickness benefits, Top 20
districts, August 2003
no. as % of total working
age population
1. Easington 8,100 14.5
2. Merthyr Tydfil 4,100 12.0
3. Blaenau Gwent 4,700 11.3
4. Glasgow 41,400 11.2
5. Knowsley 9,900 10.8
6. Neath Port Talbot 8,300 10.4
7. Liverpool 29,000 10.4
8. Barrow in Furness 4,300 10.1
9. Barnsley 11,900 8.9
10. Caerphilly 9,000 8.7
11. Rhondda Cynon Taff 12,100 8.7
12. North Lanarkshire 17.300 8.5
13. Inverclyde 4,300 8.4
14. Halton 6,200 8.4
15. Middlesbrough 6,800 8.3
16. Stoke on Trent 12,200 8.3
17. Hartlepool 4,200 8.0
18. Blyth Valley 4,100 7.9
19. Renfrewshire 8,500 7.9
20 Bridgend 6,000 7.7
Source : Authors’ estimates based on DWP data
30
Table 6 : Estimated diversion from unemployment to sickness benefits,
principal cities, August 2003
no. as % of total working
age population
Glasgow 41,400 11.2
Liverpool 29,000 10.4
Middlesbrough 6,800 8.3
Stoke on Trent 12,200 8.3
Manchester 20,600 7.5
Sunderland 12,200 7.0
Swansea 9,300 6.9
Dundee 5,100 5.7
Newcastle upon Tyne 9,000 5.4
Hull 7,600 5.1
Nottingham 8,400 4.8
Birmingham 27,800 4.6
Coventry 8,300 4.4
Norwich 3,500 4.4
Plymouth 6,400 4.3
Leicester 6,900 3.9
Bradford 10,800 3.8
Bristol 9,400 3.8
Cardiff 7,200 3.7
Derby 4,900 3.6
Aberdeen 4,800 3.5
Sheffield 10,500 3.3
Edinburgh 8,400 2.8
Brighton 4,200 2.6
Southampton 3,300 2.3
London 103,000 2.1
Leeds 9,200 2.0
Portsmouth 2,300 1.9
Source : Authors’ estimates based on DWP data
31
Figure 1: Claimants of working age incapacitated by long-term* sickness and
invalidity, 1981-2003, Great Britain
Males
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
Females
0
200,000
400,000
600,000
800,000
1,000,000
1,200,000
1,400,000
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
*for six months or more
Source: Social Security Statistics, Department of Work and Pensions
32
Figure 2: Adults of working age claiming sickness-related benefits,
England and Wales, August 2003
% of working age population
8
4
12 or over
8 to 12
4 to
0 to
Data sources: DWP and ONS Mid-year population estimates
Digital Boundary Source: Geoplan
33
34
Figure 3: Adults of working age claiming sickness-related benefits, Scotland,
August 2003
% of working age population
12 or over
8 to 12
4 to 8
0 to 4
Data Sources: DWP and ONS Mid-year population estimates
and authors' estimates
Digital Boundary Source: Geoplan
Figure 4: Hidden unemployment amongst sickness-related benefit claimants,
England and Wales, August 2003
% of working age population
or over
5
0
0
7.5 to 7.5
2.5 to 5
to 2.5
Data sources: DWP and ONS Mid-year population estimates
Digital Boundary Source: Geoplan
35
36
Figure 5: Hidden unemployment amongst sickness-related claimants, Scotland,
August 2003
% of working age population
7.5 or over
5 to 7.5
2.5 to 5
0 to 2.5
0
Data Sources: DWP and ONS Mid-year population estimates
and authors' estimates
Digital Boundary Source: Geoplan
